Using On-the-Fly Simulation for Estimating the Turnaround Time on Non-dedicated Clusters
The computation capacity of the workstations of an open laboratory in almost every university is enough to execute not only the local workload but some distributed computation. Unfortunately, the local workload introduces a big uncertainty into the predictability of the system, which hinders the applicability of the job scheduling strategies.
In this work, we introduce into our job scheduling system, termed CISNE, a simulator, which allows its scheduling decisions to be enhanced by estimating the future cluster state. This process of estimation is backed by analytic procedures which are also described in this study. Likewise, the simulation let us assure some limit to the turnaround time for the parallel user. This paper analyses the performance of the simulation process in relation to different scheduling policies. These results reveal that those policies that respect an FCFS order for the waiting jobs are more predictable than those that alter the job ordering, like Backfilling.
KeywordsSchedule Policy Cluster State Turnaround Time Parallel Application Local Load
Unable to display preview. Download preview PDF.
- 1.Acharya, A., Setia, S.: Availability and utility of idle memory in workstation clusters. In: Proceedings of the ACM SIGM/PERF 1999, pp. 35–46 (1999)Google Scholar
- 2.Litzkow, M., Livny, M., Mutka, M.: Condor- a hunter of idle workstations. In: 8th Int’l Conference of Distributed Computing Systems (1988)Google Scholar
- 10.Smith, W., Wong, P.: Resource selection using execution and queue wait time predictions. NAS Technical Reports (2002)Google Scholar
- 11.Li, H., Groep, D., Templon, J., Wolters, L.: Predicting job start times on clusters. In: 4th IEEE/ACM International Symposium on Cluster Computing and the Grid (CCGrid 2004) (2004)Google Scholar